Gut: first published as 10.1136/gut.19.1.60 on 1 January 1978. Downloaded from

Gut, 1978, 19, 60-63

Effect of bran particle size on stool weight

A. J. M. BRODRIBB AND CHRIS GROVES From the Radcliffe Infirmary, Oxford, and the Oxford Polytechnic, Oxford

SUMMARY The increase in stool weight after feeding 20 g (dry weight) of bran daily was measured when this was oflarge particle size and after milling to small particle size. Twenty-eight investigations were carried out in 21 normal subjects. With coarse bran, stool weight was significantly greater than with the fine bran (mean 219-4 g/d coarse bran, 199-0 g/d fine bran: difference 20-4 g SE + 6-4, P < 0.01). The coarse bran also had a greater water-holding capacity (7.3 g water/g coarse bran, 3.9 g water/g fine bran). Coarse bran was 21 times the volume of fine bran for a given weight and more fine bran will, therefore, be taken when bran is prescribed by the spoonful.

Wheat bran is now consumed in large quantities as was then eaten for two weeks, 20 g dry weight daily, a concentrated form of dietary fibre supplement. and all stool specimens weighed during the second Different varieties are available varying in pala- week. The other variety of bran was then eaten for tability, fibre content, and particle size (Southgate, two weeks and stool weight measured for the second 1976a, b). It has been suggested that particle size is week. Finally, no bran was eaten for the sixth and important in determining the therapeutic effective- seventh weeks and stool weight was measured for ness of bran (Kirwan et al., 1974). The finer it is, the the seventh week. Subjects were randomly allocated less its water-holding capacity (McConnell et al., to eat coarse or fine bran first and were asked to 1974), and the more susceptible in may be to maintain a similar diet throughout the seven week bacterial digestion within the colon (Nutritional period, without changes in fibre intake apart from http://gut.bmj.com/ Reviews, 1975). A study was, therefore, carried the bran. out to determine whether a change in particle size Dietary fibre, hemicellulose, cellulose, lignin, alone affects the laxative properties of bran. crude fibre, and neutral detergent fibre values for each bran were obtained (Southgate, 1976b). Method Particle size was determined by a standardised sieving test in which bran was passed through a Twenty-one volunteers, nine men and 12 women, series of screens of different mesh size and the weight on September 25, 2021 by guest. Protected copyright. measured their daily stool weight before, during, retained on each was measured. Water-holding and after eating coarse and fine bran. This was capacity was determined by the method described carried out a total of 28 times, as several subjects by McConnell et al. (1974), in which the weight of repeated the study. The volunteers, aged 20 to 40 water held by the bran after centrifugation was years, were doctors, science lecturers, nurses, measured. paramedical staff, technicians, and medical students. A large quantity of extremely coarse bran was Results carefully blended and then half of this was hammer- milled very finely with no further extraction. Both The fibre content, particle size and water-holding varieties were sterilised in dry heat at 160°C for one capacity of the two types of bran are shown in hour and were subsequently sterile on culture. The Table 1. water content of each bran was determined, and the It will be noted that, while there are only minimal bran divided into portions equivalent to 20 g dry differences in fibre content, there are considerable weight in self-sealing plastic bags. differences in particle size and water-holding capacity A normal diet was eaten for the first week of the between the two types of bran. study and all stool specimens were weighed using The basal daily stool weight was calculated as the spring balances which had previously been checked mean of 14 days (first and seventh weeks). The as accurate to the nearest gram. One type of bran increase in daily stool weight was calculated as the difference between this and the mean daily stool Received for publication 14 June 1977 weight for weeks 3 or 5. The statistical significance 60 Gut: first published as 10.1136/gut.19.1.60 on 1 January 1978. Downloaded from 61 Effect ofbran particle size on stool weight Table 1 Properties ofcoarse andfine bran The number of bowel actions per week increased with bran from a mean of 7T8 to 10-1 on coarse bran Coarse Fine was no bran and 9-7 on fine bran. There significant bran (g) (g) difference in defaecation rate between the two types Fibre content per 100 g dry wt. of bran. Those with smaller basal stool weights had Crudefibre 11-8 12-1 increase in stool weight after bran, Neutral detergent fibre 47-7 48-8 a slightly smaller Total dietary fibre 49 3 46-7 but the difference was much less than that of the Non-cellulose polysaccharide 30 4 29-9 basal weights. Cellulose (as glucose) 13-0 119 Liainin 5 9 49 There was no statistical difference in response to Particle size distribution bran between those who ate the fine and those who % by weight ofeach fraction (sieve aperture in microns) >1500 82-8 5-6 ate the coarse bran first. 1500/710 12-7 13-6 710/500 1-5 22-9 500/210 0*5 31-9 Discussion <210 25 26-0 Water-holding capacity The two types of bran were taken from the same Wt. ofwater held by 1 g bran (dry weight) 7-3 3 9 pooled sample and it was, therefore, not surprising that their fibre content was similar. The small differences were almost certainly attributable to It was more difficult to finebranbran sampling problems. get good Table 2 Laxative effect ofcoarse andjfine replicates with the coarse bran, possibly because of Bran eaten slight fractionation of the sample. The method for Mean daily stool weight (g) first determining water-holding capacity was not entirely Basal Increase with Increasewith satisfactory, as the samples did not form coherent coarse bran fine bra pellets on centrifugation, but agreement between 47 6 38 Fine duplicates was good. The higher value for the coarse 68 98 43 Coarse 80 54 34 Coare material is probably partly due to water held 81 78 72 Coarse between individual fragments as opposed to water 88 68 34 90 104 63 Faine held within the structure of the fragment. The finer 93 42 62 Fine sample would be expected to drain more thoroughly http://gut.bmj.com/ 102 82 21 Fine test The present results are 117 109 162 Coarse under the conditions. 120 78 45 Fine comparable with those of Weinreich (1976). 120 44 63 Fine The water content of the bran was variable, 129 47 44 Fine 133 15 37 Coarse especially after sterilising, and all measurementswere, 140 85 54 Coarse therefore, done on a dry weight basis. The water 141 87 56 142 154 40 Fine content of each batch was monitored and the bran 146 87 89 Coarse kept in water-tight containers until it had been on September 25, 2021 by guest. Protected copyright. 146 74 25 as water was absorbed from 147 56 49 FCionae weighed out, readily 155 68 25 Fine the atmosphere. 174 117 68 Coarse Cereal fibre is normally eaten after heat treatment 185 72 39 Coarse 187 125 78 Coarse as bread, cake, biscuits, or breakfast cereals, so this 195 84 73 Coarse was not considered detrimental to the value of 202 122 101 Coarse of cooking 204 63 70 Coarse the study, although the biological effect 219 63 51 Fine cereal fibre has not been clearly defined. 272 139 124 Fine The subjects were allowed to eat their normal Mean: 140-1 79 3 58 9 diet throughout the study and carefully avoided changes in fibre intake apart from the additional bran. The considerable difference in basal stool of the difference between the meari increase in stool weight may reflect differences in the amount of fibre weight on coarse and fine bran vwas calculated by consumed from one individual to another. Other Student's t test. The results are tabulated in Table 2, studies in progress suggest that this can vary from with the lowest basal stool weilght first and the less than 3 g to more than 48 g dietary fibre daily. largest last. Diet is often different at the weekend compared The increase in mean stool weDight with coarse with the working week, so stool weights were bran was 79-3 g and with fine bran 58*9 g. The mean measured for a complete week to allow for such difference in increase between the two was 20-4 g, normal fluctuations. Weighing was not difficult SE ± 6-4, p < 0-01 (t = 3-17). with the use of accurate spring balances and was Gut: first published as 10.1136/gut.19.1.60 on 1 January 1978. Downloaded from

62 A. J. M. Brodribb and Chris Groves found to be a convenient method that allowed for faeces (Cowgill and Anderson, 1932; Wyman et al., immediate disposal of the specimen. Nearly all 1976), and in vitro water-holding capacity tests may subjects had received a scientific training and were not be relevant to intracolonic behaviour. Coarse used to making precise measurements and keeping bran particles are probably less readily digested by accurate records. bacteria and this may result in a higher fibre content The fine bran was much more palatable than the of the faeces. Williams and Olmstedt (1936) con- coarse, and most subjects found the latter a struggle. cluded, however, in their classical study that stool Both types were consumed with breakfast cereals, weight was probably greater when more fibre was fruit, or soups mixed with the normal diet. Both digested within the colon. Thirdly, the larger tended to cause sensations of fullness, borborygmi, particles may trap more finely dispersed gas, and wind. The stools became soft, solid, of uniform produced by the colonic bacteria, increasing stool consistency and colour, were more bulky, and often bulk (Walker, 1947) and thereby increasing transit floated. time, with a consequent increase in stool weight. Table 2 shows that there was considerable in- Several subjects noticed that the coarse bran resulted dividual variation both in basal stool weights and in stools containing more gas bubbles, which in the response to the two different types of bran. floated. All three mechanisms may contribute to the This is probably due, not only to differences in diet difference between the coarse and fine bran. and transit time, but also to variations in bacterial Bran is usually prescribed by volume. One spoonful flora and innate differences in colonic function. of fine bran was equivalent in weight to 24 spoonfuls There was much less variation between studies of coarse bran. A larger weight of fine bran will, repeated with the same individual. therefore, be consumed when measured volumet- Coarse and fine brans have previously been rically. Similarly, if bran is given ad libitum more of compared by Fantus et al. (1941). They detected no the fine bran will be eaten, as it is more palatable. difference in stool weight, but different subjects ate There is no evidence of a therapeutic difference coarse and fine bran, the period of bran eating and between coarse and fine bran, provided that the stool collection for some was only three days, and amount taken results in the same increase in stool the dietary fibre content of each type of bran was weight. not shown to be comparable. Kirwan et al. (1974) studied the effectiveness of We wish to thank The Energen Foods Company, http://gut.bmj.com/ two types of commercially available bran in reducing Ashford, Kent, for the supply of bran; Dr D. A. T. colonic intraluminal pressure and transit time and Southgate, Dunn Nutrition Laboratory, Cambridge, found that fine bran was less effective than coarse for the dietary fibre analysis; Dr D. J. Stevens, bran after four weeks' treatment. The water-holding Flour Milling and Baking Research Association, capacity of the coarse bran was 6-15 g water per g Chorley Wood, for crude fibre, neutral detergent bran and 2 36 g water per g for the fine bran. They fibre, particle size and water-holding capacity concluded that the particle size ofbran was important measurements; Dr D. Quantrell, Radcliffe Infirmary, on September 25, 2021 by guest. Protected copyright. in determining stool bulk and therapeutic efficacy. for bran water content measurements; and parti- This conclusion has been criticised because the two cularly the volunteers for their help. groups were not closely matched and the amount of fibre present in each bran was different, patients on coarse bran receiving over 3 0 g acid detergent fibre References daily compared with 1 9 g on fine bran. Cowgill, G. R., and Anderson, W. E. (1932). Laxative effects The present study shows that particle size is of wheat bran and 'washed bran' in healthy men. Journal of important in determining the effect of bran in American Medical Association, 98, 1866-1875. increasing stool weight. It can no longer be assumed Fantus, B., Hirschberg, N., and Frankl, W. (1941). The mode of action of bran. II. Influence of size and shape of bran that fibre of the same chemical composition has the particles and of crude fiber isolated from bran. Review of same biological or therapeutic effect on the colon, Gastroenterology, 8, 277-280. even if from the same source and consumed in Kirwan, W. O., Smith, A. N., McConnell, A. A., Mitchell, identical quantities. This might explain some of the W. D., and Eastwood, M. A. (1974). Action of different bran preparations on colonic function. British Medical inconsistencies between different clinical trials. Journal, 4, 187-189. There are at least three different ways in which McConnell, A. A., Eastwood, M. A., and Mitchell, W. D. particle size could influence stool weight. The larger (1974). Physical characteristics of vegetable foodstuffs that particles have a higher water-holding capacity and could influence bowel function. Journal of the Science of Food and Agriculture, 25. 1457-1464. could, therefore, increase the relative amount of Nutrition Reviews (1975). Dietary fiber and colonic function- water present in the stool. In fact, bran does not an effect of particle size? Nutrition Reviews, 33, 70-72. greatly increase the percentage water content of the Southgate, D. A. T. (1976a). Dietary fibre (Letter). British Gut: first published as 10.1136/gut.19.1.60 on 1 January 1978. Downloaded from

Effect ofbran particle size on stool weight 63 Medical Journal, 2, 236. Williams, R. D., and Olmstedt, W. H. (1936). The effect of Southgate, D. A. T. (1976b). Analysis of dietary fibre. In cellulose, hemicellulose, and lignin on the weight of the Fiber in Human Nutrition. Edited by G. A. Spiller, and stool: a contribution to the study of laxation in man. R. J. Amen. Plenum Press: New York. Journal of Nutrition, 11, 433-449. Walker, A. R. P. (1947). The effect of recent changes of food Wyman, J. B., Heaton, K. W., Manning, A. P., and Wicks, habits on bowel motility. South African Medical Journal, A. C. B. (1976). The effect on intestinal transit and the 21,590-596. feces of raw and cooked bran in different doses. A4merican Weinreich, J. (1976). Bran and the irritable bowel (Letter). Journal of Clinical Nutrition, 29, 1474-1479. Lancet, 1, 810-811.

The December 1977 Issue THE DECEMBER 1977 ISSUE CONTAINS THE FOLLOWING PAPERS IgA and IgG reticulin antibodies in coeliac and Clinical study ofliver blood flow in man measured by non-coeliac patients 0. E. EADE, R. S. LLOYD, CELIA 133Xe clearance after portal vein injection s. B. LANG, AND R. WRIGHT SHERRIFF, R. C. SMART, AND I. TAYLOR Chronic laxative abuse associated with pancreatic Localisation of e-antigen in nuclei of hepatocytes in islet cell hyperplasia M. LESNA, A. N. HAMLYN, HB%Ag-positive liver diseases w. ARNOLD, J. O. C. W. VENABLES, AND C. 0. RECORD NIELSEN, F. HARDT, AND K. H. MEYER ZUM BUSCHEN- FELDE Qualitative and quantitative evaluation of internal anal sphincter function in the newborn V. E. Immunofluorescence detection of new antigen- BOSTON, S. CYWES, AND M. R. Q. DAVIES http://gut.bmj.com/ antibody system (8/anti-8) associated to hepatitis B Policy of local excision for early cancer of the virus in liver and in serum of HBsAg carriers colorectum B. C. MORSON, H. J. R. BUSSEY, AND M. RIZZETTO, M. G. CANESE, S. ARICO, 0. CRIVELLI, S. SAMOORIAN C. TREPO, F. BONINO, AND G. VERME Giant sigmoid diverticulum: clinical and radiologi- cal features D. R. FOSTER AND B. ROSS

T-lymphocyte cytotoxicity to HBsAg-coated target on September 25, 2021 by guest. Protected copyright. cells in hepatitis B virus infection A. ALBERTI, Changes in the plasma clearance and protein binding G. REALDI, F. BORTOLOTTI, AND A. M. RIGOLI of carbenoxolone with age, and their possible relationship with adverse drug effects M. J. HAYES, K-lymphocytes (killer-cells) in Crohn's disease and MARGARET SPRACKLING, AND M. J. S. LANGMAN acute virus B-hepatitis R. ECKHARDT, P. KLOOS, M. P. DIERICH, AND K. H. MEYER ZUM BUJSCHENFELDE Short report Failure of glucagon to influence ion transport across human jejunal and ileal mucosa in vitro P. E. T. ISAACS AND L. A. TURNBERG Reduction in peripheral blood K cells and activated T cells in primary biliary cirrhosis G. P. SANDILANDS, Progress report Transplanation of insulin-secreting R. N. M. MACSWEEN, KATHLEEN G. GRAY, R. J. HOLDEN, tissues R. C. KARL, D. W. SCHARP, W. F. BALLINGER, P. MILLS, FIONA M. REID, MAIR A. THOMAS, AND AND P. E. LACY G. WATKINSON Notes and activities Pathophysiology of hypotension in patients with Books fulminant hepatic failure P. N. TREWBY AND ROGER WILLIAMS Index to Volume XVIII Copies are still available and may be obtained from the PUBLISHING MANAGER, BRITISH MEDICAL ASSOCIATION, TAVISTOCK SQUARE, LONDON WC1H 9JR, price £2-75, including postage